Fixed costs, by definition, are costs that do not depend on the level of production. However, just like in the case of variable costs, this is not that simple. Yes, a company may pay year over year about the same for its corporate headquarters and its staff, but if production (and revenue) grows fast, such outlays tend to become larger.
So, in effect, fixed costs are not that fixed after all. What is important to know is that they are not as adaptable to the level of production as variable costs are, and that if they change substantially, they tend to change in a step-up (or step-down) manner, i.e. not gradually. Their change, though affected by changes in the production volume, which tend to change gradually (together with demand), is really set by a corporate decision, and as such is not that predictable.
You, our reader, might be puzzled by now: why are we talking about variable and fixed costs if the company itself might not keep track of them and they are definitely not shown as such on any company income statement? Well, if we want to make a reasonably good prediction about the company's future cash flows (and this is usually the objective for financial statements analysis done with the purpose of business or investment valuation), we'd better know how total costs change with a change in revenue. If we can split, even approximately, total operating costs into variable and fixed costs, our prediction capability increases dramatically. This also helps in understanding the business and key factors affecting its profitability.
Though splitting of total operating costs into variable and fixed costs is a mandatory feature of any microeconomic textbook, its not a common practice at all in financial analysis, as the splitting of expenses into variable and fixed is not that straightforward.
Well, as it was already said, investment valuation is more of an art than science. Everybody practices it in a slightly (or sometimes totally) different way. Our approach is decidedly different from others, though. And one of its distinguishing features is a separate forecasting of fixed and variable expenses.
So, how do we go about splitting total operating expenses into fixed and variable? For this we analyze how total expenses changed in relation to revenue in the past and use a simple formula to calculate the split. The formula and the process how it was derived is described in the chapter "Chepakovich valuation model" below.
Our cost splitting methodology relies on regular financial statements produced by the company and, therefore, does not always yields a meaningful result (financial statements just do not explicitly contain the information we are after). Nonetheless, just like in many other areas, having an instrument that works at least half of the time is better than having no instrument at all.
Once you are finished with forecasting revenue and expenses, half of the work of building a cash flow model is done. The rest is a lot easier and straightforward, and, we might add, more rigorous.